High-Level Overview
NetworkOcean builds and operates underwater data centers that leverage seawater cooling to reduce power usage by up to 30-40% and eliminate freshwater consumption entirely, creating more sustainable and capital-efficient AI infrastructure[2][3]. Their product includes modular underwater data center capsules designed for efficient cooling and high-density GPU deployment, currently offering reservations for 2,048 Nvidia H100 GPUs[2]. They primarily serve AI companies and cloud infrastructure clients seeking lower-cost, energy-efficient, and environmentally friendly data center solutions[2]. By addressing the critical challenges of power and water consumption in data centers, NetworkOcean aims to reduce operational costs and environmental impact while supporting the growing demand for AI compute capacity.
Origin Story
Founded by Sam Mendel and Eric Kim, NetworkOcean emerged from their shared background in robotics and ocean technology, dating back to high school[2]. Both founders have strong technical and entrepreneurial experience: Eric holds patents in renewable energy and has worked on computer vision and large language model applications, while Sam has built multiple startups and hardware projects, including a floating web server deployed in San Francisco Bay[2]. The idea to build underwater data centers grew from their expertise in underwater systems and a vision to solve the energy and water challenges of traditional data centers. The company is backed by Y Combinator and is preparing to test a 0.5 MW underwater data center capsule in San Francisco Bay[1][2].
Core Differentiators
- Energy Efficiency: Uses seawater cooling to reduce power usage by up to 40%, significantly lowering operational costs and carbon footprint[3].
- Water Conservation: Eliminates freshwater consumption, addressing a major environmental concern for hyperscale data centers that rely on water-intensive evaporative cooling[2][3].
- Modular Design: Deploys underwater data center capsules that can be clustered, enabling scalable and flexible infrastructure[1][2].
- GPU-Optimized: Offers high-density GPU capacity (2,048 Nvidia H100 GPUs available) tailored for AI workloads[2].
- Rapid Deployment: Claims deployment times as short as four months, faster than many traditional data center builds[3].
- Founders’ Expertise: Deep technical background in underwater technology and AI infrastructure, enabling innovative design and operational strategies[2][4].
Role in the Broader Tech Landscape
NetworkOcean is riding the trend toward sustainable and efficient AI infrastructure, addressing the growing energy and water constraints faced by hyperscale data centers amid rising AI compute demand[2]. The timing is critical as traditional data centers face scrutiny over water use and energy consumption, with regulatory challenges emerging globally (e.g., Google’s permit issues in Chile)[2]. Underwater data centers offer a promising alternative by leveraging natural seawater cooling and reducing land use, which aligns with increasing corporate and regulatory pressure for greener technology solutions. NetworkOcean’s approach also builds on earlier experiments by Microsoft and others but aims to move beyond proof-of-concept to scalable, operational deployments[1][2]. Their work could influence the broader ecosystem by setting new standards for sustainable data center design and enabling more cost-effective AI infrastructure.
Quick Take & Future Outlook
NetworkOcean is positioned to capitalize on the urgent need for sustainable AI infrastructure by scaling underwater data centers that cut energy and water use significantly. The upcoming test of their 0.5 MW capsule in San Francisco Bay will be a pivotal milestone demonstrating operational viability. Future trends shaping their journey include increasing AI compute demand, tightening environmental regulations, and the push for carbon-neutral technology. As they expand, NetworkOcean could become a key player in redefining data center architecture, influencing both the cloud infrastructure market and environmental sustainability standards. Their success could accelerate adoption of ocean-based data centers globally, especially in coastal regions with high compute demand and water scarcity concerns.
